Content uploaded by Gonçal Costa
Author content
All content in this area was uploaded by Gonçal Costa on Sep 10, 2016
Content may be subject to copyright.
Content uploaded by Gonçal Costa
Author content
All content in this area was uploaded by Gonçal Costa on Sep 10, 2016
Content may be subject to copyright.
1 INTRODUCTION
Energy efficiency is a fundamental part of Europe
2020 strategy to reduce the energy footprint and CO2
emissions related to new and renovated buildings.
To help achieve this aim, the optimized, integrated
and systemic design of energy efficient retrofitted
buildings is also fostered at district scale. To be
more efficient in this design, new technological solu-
tions to carry out different types of simulations
through comprehensive tools are necessary to assess
the impact of different actions through a holistic ap-
proach. In this approach, these tools should be able
to integrate the different data required to perform
these simulations including GIS, BIM, energy, eco-
nomic, weather, monitoring, social, targets, etc.
One way to assess the impact on energy efficiency
of buildings is by applying energy conservation
measures according to the needs of the project and
compare the results. On the basis of this approach
has been devised an “Optimised Energy Efficient
Design Platform for refurbishment at district level”
in the context of OptEEmAL project
1
. This paper
presents the project’s vision, lines and general ap-
proach proposed to support the retrofitting design in
the simulation terms described above.
The platform aims to help in decision-making in
retrofitting projects at district level. There are sever-
1
https://www.opteemal-project.eu/
al factors which can play a role in the decision-
making process and explain the reasons why a dis-
trict can be retrofitted, for example, those related to
improve the quality of the inhabitants, the use of
spaces, the economic dynamism of the areas, or re-
duce pockets of poverty, among other causes. These
factors are not only related to energy consumption,
but also with social or economic aspects to be con-
sidered to reach cost-effective solutions and more
sustainable developments. Based on these factors,
different actions can be taken in the definition of a
district retrofitting plan. However, this often requires
stakeholders to make complex decisions through the
use of different tools and where the objective param-
eters have to be brought to light and investigated
through specific technical studies. To overcome
these drawbacks and to better support the design of
district retrofitting plans and strategies through a ho-
listic approach, solutions able to integrate stakehold-
ers, tools and all relevant information about retrofit-
ting in a single framework; such solutions seem to
be missing.
The management and conservation of urban dis-
tricts requires an approach that considers each of the
buildings and other city elements as forming part of
an environment that should be conserved, brought
up-to-date and showcased. This approach requires
the integration of Geographic Information Systems
(GIS) and Building Information Models (BIM),
A comprehensive ontologies-based framework to support retrofitting
design of energy-efficient districts
G. Costa & Á. Sicilia
ARC, La Salle Engineering and Architecture, Ramon Llull University, Barcelona, Spain
G.N. Lilis & D.V. Rovas
Department of Production Engineering and Management, Technical University of Crete, Chania, Greece
J. Izkara
Construction unit, Tecnalia Research & Innovation, Bizkaia, Spain
ABSTRACT: One of the challenges for the European construction sector as a part of the Europe 2020 strategy
is to reduce the energy footprint and CO2 emissions related to new and renovated buildings, an interest that is
also fostered at district scale requiring new technological solutions to be more efficient in this design. In re-
sponse to this challenge, a web-based platform for district energy-efficient retrofitting design projects has
been proposed in the context of OptEEmAL research project. In order to provide data integration and interop-
erability between BIM/GIS models and energy simulation tools through this platform, a District Data Model
(DDM) has been devised. This way, representations of CityGML and IFC schemes semantically enriched are
related in this model to existing ontologies in the main fields for urban sustainable regeneration (energy, so-
cial, environment, comfort, urban morphology and economic). This paper discusses some approaches about
how consistency integration of geometry and the semantic representation from IFC and CityGML files with
different levels of detail can be achieved in a district data model.
while at the same time bearing in mind the particular
nature of urban districts (Döllner & Hagedorn 2007).
On the other hand, when the retrofitting is ad-
dressed at district scale the complexity of decision
making grows exponentially due to the great number
of factors to be considered (e.g. economic, social,
technical), the interactions between them, and the
number of stakeholders involved in the decision.
Consequently, there is a need for an interactive and
user-friendly decision support tool that enables anal-
ysis of the impact of the building energy oriented ret-
rofitting project on the sustainability of the urban
district in a holistic way, facilitating the necessary
communication mechanisms that can forge agree-
ment between the multiple stakeholders that are in-
volved in this process (Romero et al. 2014).
The paper is structured as follows: Section 2 in-
troduces the OptEEmAL platform as a candidate so-
lution addressing the need to apply energy efficient
district retrofitting actions at district level. An ontol-
ogy-based approach to facilitate district data integra-
tion and promote interoperability with multiple sim-
ulation tools is introduced as the key component of
this platform. The three essential components of this
platform: (a) Energy Conservation Measures Cata-
logue, (b) District Data Model, and the (c) Simula-
tion Manager, are also introduced in Section 2. The
role of ontologies to facilitate the integration of dif-
ferent data models and tools is reviewed in Section 3
in order to provide the theoretical base to explain
how the DDM provides the intertwining of standard
data models (e.g., CityGML, IFC) with ontologies in
domains related with sustainable regeneration (ener-
gy, social, environment, comfort, urban morphology
and economic), which is described in Section 4.
Conclusions and discussion on ongoing research in
the OptEEmAL project is provided in the last section
of this paper.
2 OPTEEMAL PROJECT
2.1 Context and purpose
In order to respond to the need for innovative design
tools for refurbishment at building and district level,
a development of a platform has been proposed in
the context of OptEEmAL project with the aim of
delivering an optimized, integrated and systematic
design for building and district retrofitting projects.
Based on given initial district conditions, the plat-
form provides the necessary information to simula-
tion tools according to multiple candidate energy
conservation measures (ECM). This information in-
cludes buildings, urban areas, weather, sensors, etc.,
and project conditions (costs, barriers, targets, etc.).
Through a comparison of the respective simulation
results, the platform identifies the most suitable en-
ergy conservation measure, which achieves a desired
reduction of the district energy demand and con-
sumption under certain constraints. By selecting the
best case, the design is updated according to the
measures implemented to obtain an enhanced ver-
sion which can be returned to the BIM authoring
tools.
District and building scales have been addressed
separately but they both are very much connected in
the process of energy efficient retrofitting of dis-
tricts. Strategic decisions — such as prioritization of
areas to retrofit and implementation of district heat-
ing — are taken at district scale, while executive de-
cisions are mainly addressed at building level. Urban
scale and the influence and restrictions imposed by
urban environment in the building retrofitting should
be taken into consideration at initial stages of the ret-
rofitting process (e.g. feasibility studies and concep-
tual design). In early stages of an energy retrofitting
process, administrations and managers are mainly
involved and the level of detail of the information
required is low. In later stages of the process (e.g.
design or implementation of the interventions) a de-
tailed description of building components and their
characteristics are required. Main stakeholders in
these steps are architects and constructors, and deci-
sions to be taken are focused on the building level,
even more on component level. In this context, it is
critical to identify solutions able to cover the need
for the connection between the strategic scale (ur-
ban) and the executive scale (building), through the
definition of a common, multi-scale and interopera-
ble data model which contains geometric and seman-
tic information required for the management and de-
cision making in the district retrofitting.
2.2 Platform components
The OptEEmAL platform is composed of three main
components in order to provide services for current
situation diagnosis, retrofitting scenarios generation,
evaluation and optimization and data export. These
components are (1) an integrated ontology-based
District Data Model (DDM) that is connected to a
(2) catalogue of Energy Conservation Measures
(ECMs) and which is accessed by a (3) simulation
manager that is responsible to generate specific sim-
ulation models for each tool required (figure 1).
The DDM is the central component of the plat-
form and has been conceived as a comprehensive
ontologies-based framework for district information
representation based on the intertwining of standard
data models (e.g., CityGML, IFC) with ontologies in
domains related with sustainable regeneration (ener-
gy, social, environment, comfort, urban morphology
and economic). The DDM provides a semantically
integrated data model (including information about
the geometry, materials, equipment, and indicators,
at the building and urban scales) that the platform
needs to carry out retrofitting processes.
Energy conservation measures (ECM) both at the
building and district level, are contained in an ECM
catalogue. These measures contain key information
to generate applicable scenarios, but also to over-
coming the existing barriers in the district and being
compliant with user objectives in terms of efficiency
improvement, cost constraints, financial schemes,
etc. The catalogue includes a wide range of measures
to reduce the district energy demand and consump-
tion through passive, active, local Renewable Energy
Sources (RES) integration and control strategies
measures.
The simulation manager implements different
calculation methodologies that are necessary to gen-
erate simulation models. These models include the
necessary information to be processed by external
tools (e.g. EnergyPlus). From the results of the simu-
lation tools, District Performance Indicators (DPIs)
are computed. In summary, the simulation manager
calculates the DPIs referring to (a) the baseline sce-
nario to diagnose the current status of the district,
and (b) from the retrofitting scenarios generated by
applying different energy conservation measures.
Once the DPIs are calculated according to possible
optimized designs, they are compared with those
calculated in the baseline design in order to assess
the level of improvement in the energy performance.
Figure 1. Overview of the architecture platform outlined in
OptEEmAL project.
2.3 Performance evaluation
DPIs are defined in order to assess the impact of the
measures contained in the ECM catalogue using dif-
ferent performance criteria, which include energy
demand, consumption, cost and others. In this way
the selected scenarios not only conform to a number
of different performance constraints but also achieve
desired performance values. Some of the DPIs re-
quire complex calculations while others are obtained
by simple operations using other DPIs. DPIs requir-
ing complex operations are evaluated through build-
ing simulations. In order to evaluate these DPIs in an
automated manner and assess different retrofitting
scenarios in a relatively short time, an automated
simulation model generation process is established,
using data across different time and space scales.
2.4 Data requirements
The required data of the OptEEmAL platform can be
classified using space and time criteria. Using a spa-
tial differentiation these data can be distinguished in-
to (a) BIM and (b) district data. BIM data refer to
each building individually and contain information
regarding its geometric description, the materials of
its constructions and the passive or active devices
which may be present in its interior spaces. District
data are related to building groups and include geo-
metric description of multiple building envelopes,
and district level systems, serving multiple build-
ings, such as district heating.
Using time criteria, the required data can be clas-
sified into: (a) static data, which remain unchanged
during simulation executions such as building geo-
metric data, and construction material data, and (b)
dynamic data which change during simulation exe-
cutions such as the operation schedules of building
devices.
Managing the above plethora of different data ap-
pears to be a challenging issue for the automation of
the simulation model generation process performed
by the simulation manager. This automation involves
careful selection of subsets of the above data sets
based on the requirements and the characteristics of
each individual simulation. Two general simulation
cases can be distinguished: 1. District-scale simula-
tions performed by CitySim where the district data
are required, and 2. Building-scale simulations per-
formed by EnergyPlus where both district and build-
ing-scale data must be taken into account. To cover
both simulation types IFC, CityGML and Contextual
data are required, as described next.
2.5 Input data sources: IFC, CityGML and
Contextual data
Some of the data required for the population of
DDM are coming from a variety of existing data
structures such as the widely used IFC and CityGML
schemas and some —not contained in the previous
schemas— are inserted manually, characterized as
contextual data. IFC is a popular data schema adopt-
ed by the AEC industry as a standard (ISO 16739
2013) to describe a variety of building entities in-
cluding architectural, structural and mechanical
components. In the context of the present work data
from the architectural section of IFC are required to
describe the geometry of the buildings and the prop-
erties of building materials and information from the
mechanical part will be used to describe the installed
systems in the buildings.
District-level data are collected from CityGML
data structures. These include mostly geometric data
of the building envelopes and also the geometric rep-
resentation of other urban elements (e.g. green areas,
roads, city furniture). These other urban element da-
ta, will play an indirect role in building simulations
as shading surfaces and will not be a part of any ret-
rofitting scenario. These data will either augment
each individual building simulation models in order
to perform better shading calculations by including
additional neighbor shading objects or will provide
the necessary geometric input of district-scale simu-
lation models suitable for CitySim.
IFC and CityGML structures do not provide all
the necessary data for simulation model generation.
Missing data appear in both building and district
contexts and are characterized as contextual data.
These include: weather data, operation schedules of
devices and inhabitants, simulation parameters, en-
ergy prices and building typologies (Vimmr et al.
2013).
Each individual data component before being in-
serted in the DDM should pass three checking stag-
es: correctness, completeness and consistency. Dur-
ing the correction stage the inserted data are checked
for compliance to certain correctness rules. Towards
this direction, error detection and correction mecha-
nisms, such as the ones developed for the geometric
data of IFC files (Lilis et al. 2015), can be used in
order to guarantee data correctness in DDM. Simi-
larly, during the completeness stage, the data insert-
ed in the DDM are checked against certain com-
pleteness rules. For example, completeness rules are
defined by the minimum data requirements for simu-
lation model generation. In case any of these re-
quirements is not satisfied, a completeness error is
reported. Finally, at the consistency stage, the insert-
ed data structures are checked for compatibility to
other existing DDM data structures. For example, an
inserted IFC BIM model should be correctly placed
(location/orientation) with respect to an existing
three dimensional CityGML geometric context. Any
inconsistency is also reported for correction.
3 USE OF ONTOLOGIES TO INTEGRATE
DATA MODELS AND TOOLS
3.1 Implementation scenarios for data integration
Several European initiatives and guidelines consider
data models of vital importance for improving the
energy performance of buildings information (e.g.
EeB
2
, ECTP
3
) and try to establish a common geospa-
tial information infrastructure at European Level
(e.g. INSPIRE
4
), coming to the conclusion that a bet-
ter understanding of the urban system is necessary in
order to achieve sustainable development goals for
cities (e.g. EPIC
5
, SEMCITY
6
). Accurate 3D urban
models are an important tool for a better understand-
ing of urban systems and thus for sustainable urban
development. The solution, based on 3D digital
models, has grown in importance over recent years
as it offers complete support which is easily brought
up-to-date, allowing information storage and visuali-
zation on an urban scale (Mao & Ban 2011).
There are international standards for the man-
agement of data related with construction processes
based on BIM (Building Information Modelling) and
GIS (Geospatial Information System). At building
level, data models based on XML facilitate the vali-
dation and exchange between Computer-Aided De-
sign applications or energy assessment tools
(gbXML, Architecture Engineering and Construction
XML, Building Information Model XML, Industry
Foundation Classes XML, etc.). At urban or city
scale GML and KML are the most used data formats
for 3D representation. However both can store ge-
ometry but are not designed to store semantic infor-
mation. At building scale, the problems are similar,
CAD tools used to work with lines and polygons and
the semantic information available was almost zero.
When the BIM concept appeared and Industry Foun-
dation Classes (IFC) implemented the international
open standard for BIM (Succar 2009), a significant
step forward has been made in the semantization of
the building scale model. A multi-scale data model
that integrates both scales is CityGML. The aim of
the development of CityGML was to reach a com-
mon definition and understanding of the basic enti-
ties, attributes, and relations within a 3D city model
(Kolbe 2009). What is especially important, since it
allows the reuse of the same data in different fields
of application (Gröger & Plümer 2012). It has been
designed to store both types of information, allows
storing 3D information, considering both urban scale
and building level. CityGML is a standard widely
used in Europe (most of the German cities have a
CityGML model at least at its lowest level of detail
(LoD1), some of them with highest level. Berlin has
one of the most advanced CityGML model, Dutch
3D standard IMGeo is a CityGML ADE, etc.).
Combining domain specific information with data
models for the construction sector is a task being ad-
dressed through different approaches. Urban energy
2
http://ec.europa.eu/research/industrial_technologies/energy
-efficient-buildings_en.html
3
http://www.ectp.org
4
http://inspire.ec.europa.eu
5
http://www.semcity.net/cms
6
http://www.epic-cities.eu
tool developers at city level (e.g. CitySim, NEST)
have developed their own tailor made urban infor-
mation. Extension of existing standard data models
for construction sector (e.g. IFC, CityGML) in order
to complete them with domain specific information
has been mainly addressed through the use of exten-
sion mechanism defined for existing data models
(e.g. ADEs for CityGML). This approach allows
storing relevant domain specific data in a common
open city data model, used to perform domain spe-
cific simulation. However, the interoperability of the
data model is reduced when the extension is imple-
mented and the extension is tailored to the data
model. The use of ontologies and vocabularies for
the conceptualization of scenarios allows represent-
ing data from different domains in order to carry out
holistic analysis. Multiple domains, scales, and lev-
els of detail have to be modelled such as urban reno-
vation projects (Crapo et al. 2011, Sicilia et al.,
2014), this is particularly important in the necessary
definition of scenarios.
3.2 Semantic-based interoperability
Semantic interoperability solutions are based on
providing a shared understanding of the meaning as-
sociated to the data from different sources and do-
mains in order to facilitate the exchange across net-
worked information systems. The meaning can be
provided through ontologies and making explicit the
semantics of data through formal languages. Ontolo-
gies specified in OWL (Ontology Web Language) al-
low the specification of a description logic based
formal structure to RDF graphs. If an ontology in
OWL is instantiated in an RDF graph, generic que-
ries and reasoning engines are able to easily reuse
the data of this graph.
Semantic-based solutions can be applied to inte-
grate data from different data models including IFC,
CityGML and data from other sources such as cadas-
tre, climate, consumption of buildings, and others.
Semantic-based interoperability using Semantic Web
technologies is a reasonable technological solution
to integrate data from multiple heterogeneous
sources and to ensure the communication between
the integrated data and an open set of tools. The use
of semantic technologies to enhance IFC and
CityGML has already been explored in some re-
search works (Laat & Berlo 2011, Amirebrahimi et
al. 2015, and others).
3.3 Data integration and transformation process in
OptEEmAL
The goal of the data integration and transformation
process in the OptEEmAL platform is to populate
the District Data Model with different input data
models provided by an end-user (Figure 2). In the
first step of the process, the information from input
data models is transformed into semantic data mod-
els by means of ontologies which define the particu-
lar domain of the input data. Then, in between the
semantic data models and the simulation tools, there
are the simulation data models which are ontology-
based models to represent a simulation domain such
as energy and economic. The simulation data models
are generic enough and are representative in order to
feed different simulation tools. From the simulation
data models are derived the final simulation models
which are particular from each simulation tool. For
example, the simulation model of EnergyPlus is the
input data files (IDF). For NEST (Neighborhood
Evaluation for Sustainable Territories)
7
, which is a
tool for Life Cycle Assessment (LCA) calculation,
the input data file is a proprietary format. For City-
SIM
8
, which is an urban performance simulation en-
gine, the input data file is a XML.
In the DDM, the semantic data models and simu-
lation data models must be represented by ontologies
that define the particular domain of the models. This
is needed (1) to carry out the transformation of input
data to a specific semantic domain, and (2) to pro-
vide a representation in RDF format in order to facil-
itate their querying through SPARQL. Queries in
this language enable to retrieve data to generate sim-
ulation models in a flexible way.
Since ontologies are required to represent the in-
put data (IFC, CityGML and contextual data) in
RDF in order to facilitate their integration into simu-
lation data models such as the EDM (Energy Data
Model).
Figure 2. Overview of the data integration and transformation
process in OptEEmAL platform. In this figure the process is
exemplified for the case of EDM, and EnergyPlus, CitySim and
NEST tools.
Different prototypes of ontologies in these do-
mains have been developed in the last decade in or-
der to provide their representation as semantic data.
For example, Katranuschkov et al. (2003) created an
ontological framework as part of an extensible and
open architecture to access data in IFC format. More
7
http://www.nobatek-nest.com
8
http://citysim.epfl.ch
recently, and as a result of some research projects
(Schevers & Drogemuller 2005, Beetz 2009, Pau-
wels & Terkaj 2016), IFC is now available as an on-
tology (ifcOWL) with the support of the Build-
ingSMART. The ifcOWL ontology enables
extensions towards other structured data sets using
semantic web technologies (buildingSMART 2015).
Regarding CityGML, Métral et al. (2010) presented
various approaches based on the use of ontologies to
improve the interoperability between 3D urban mod-
els. This helped to demonstrate that ontologies can
overcome the semantic limitations in CityGML data
models.
In the energy domain, SimModel is the prime ex-
ample of simulation data model (O’Donnell 2011). It
is an xml-based data scheme designed to support
building-scale simulation models which has an OWL
version (Pauwels, 2014). SimModel does not contain
district-related data structures. Therefore, to inte-
grate district data an extension can be developed.
Since, simulation data models — such as SimModel
— are represented by means of ontologies, they can
be easily extended.
The data integration and transformation process
in OptEEmAL is based on three steps:
ETL1: between data models and semantic data
models. This is a transformation from raw data
sources stored in CSV files, relational databases,
XML, Json, etc., to RDF. In the Semantic Web
community exists several technological solutions
to deal with this kind of sources such as relation-
al-to-RDF translators (e.g., morph-RDB) and
mapping languages (e.g., R2RML).
ETL2: between semantic data models and simula-
tion data models. This is a transformation from a
RDF graph to another RDF graph with a different
structure defined by simulation data models (e.g.,
SimModel).
Model generation: between simulation data mod-
els and simulation models. This transformation
has to be created ad-hoc for each simulation tool
using SPARQL queries.
4 DISTRICT DATA MODEL (DDM)
4.1 Data integration process to generate simulation
models
The approach adopted in the OptEEmAL platform in
order to assess the performance of district retrofitting
scenarios, through calculation of different types of
DPIs, is divided into four stages: (1) input data
quality checking, (2) transformation of input data to
OWL/RDF input data (structured according to se-
mantic data models: CityGML OWL, IFC OWL,
Contextual Data OWL and ECM data OWL), (3)
conversion of OWL/RDF input data to OWL/RDF
simulation data models (structured according to se-
mantic data models: SimModel Extended OWL and
Other Simulation Domains OWL), and (4) genera-
tion of simulation models, as illustrated in Figure 3.
This section shows the role played by the DDM in
this process to ensure the interoperability between
simulation data models and tools, and showing its
relation with the rest of the platform components
previously introduced.
The process starts with the entry of project data
by users of the platform (a CityGML model and dif-
ferent IFC models (noted as IFCb), but also other da-
ta such as socio-economic, sensors monitoring, en-
ergy prices and weather). Input data is checked in
this first part of the process to verify its correctness,
completeness and consistency checking, as described
in section 2.5. Since the information provided in the
IFC files cannot be used directly as inputs to energy
simulation programs as they require further pro-
cessing related to the generation of the second-level
space boundary geometric topology, a converter is
used to made this transformation of these IFC files
into a Boundary Surface Topology (BST) (Lilis et al.
2016).
In a second step, the input data from CityGML
and IFC files are transformed into data described in
OWL and RDF languages, according to ontologies
that correspond with versions of these standards, and
ontologies from other domains related with sustaina-
ble regeneration (energy, social, environment, com-
fort, urban morphology and economic). Through a
mapping between ontologies it is possible to perform
ETL (Extract, Transform and Load) processes in
which the data defined in each of the input domains
are transformed to different simulation data models.
When the data are defined in the domain of a simula-
tion data model, measures from the ECM Catalogue
can be applied as new data aggregated, for example,
in as new properties of elements and materials. In-
formation described in the simulation data models
(e.g. EDM) is queried through SPARQL queries to
generate the simulation models in the last step of this
process, which are generated according to each spe-
cific tool (EnergyPlus, SimCity, NEST, etc.).
The two following sections show in more detail
how the simulation models are generated and how
the type of definition of the input data provided is re-
lated to different levels of accuracy for the calcula-
tion of the DPIs using a methodology.
4.2 Simulation data models
The simulation programs which will be used for DPI
evaluation (EnergyPlus, CitySim, NEST) require dif-
ferent forms of input data populating different simu-
lation data models. Furthermore, depending on the
data availability, different calculation methodologies
can be established using the same simulation tool.
As a result, multiple simulation data models can be
formed, using different queries as part of the ETL
process mentioned earlier, depending on the selected
simulation tool and calculation methodology (as highlighted in figure 3, step 3).
Figure 3. Data flow of the process to generate simulation models.
In EnergyPlus input data must be structured using
different classes defined, depending on the version
of the program, in an input data dictionary file (*.idd
file). Based on the data class descriptions contained
in the IDD file, a single input data file (*.idf file)
will be generated by the simulation manager for each
selected retrofitting scenario and calculation meth-
odology.
Similarly in CitySim, inputs are structured in an
xml file format according to a predefined XSD
schema. Again here, this XML file will be populated
based on the selected scenario of and calculation
methodology.
Finally, the interface of OptEEmAL platform
with the NEST tool and the related NEST input file
data format, remains as a future work.is a future
work direction.
4.3 Calculation methodologies
Sophisticated simulation tools such as EnergyPlus,
accept different representations in multiple input da-
ta types (DTs). For example the building construc-
tion, input data type, may have two descriptions, a
detailed multi-layer description where the properties
of each material layer are defined, or an equivalent
single layer description. Such variability in the accu-
racy of the description of the input DTs, generate
multiple simulation execution possibilities, defined
as calculation methodologies (CMs), for every simu-
lation tool (ST). The expected accuracy of the DPI
evaluation of each CM, varies as well, as illustrated
in figure 4. Consequently for each refurbishment
scenario and simulation tool, a finite number of cal-
culation methodologies (CM) can be established,
depending on the number of possible combinations
of the input data type descriptions. Each CM is de-
fined by a unique combination of input data type de-
scriptions, highlighted by the solid arrow in figure 4.
Figure 4. Expected DPI accuracy variation depended on the se-
lected calculation methodology.
5 CONCLUSIONS
This paper has introduced a District Data Model de-
vised for the OptEEmAL platform as a framework to
support retrofitting designs in districts. Its imple-
mentation as an ontology-based approach has been
discussed in this paper as a possible solution to inte-
grate different design information (BIM, GIS and
contextual data) necessary to perform simulations
using different tools.
The use of simulation data model in the context
of the DDM is presented as a plausible approach to
solve the interoperability issues between data models
(e.g, IFC, CityGML) and different simulation tools.
In the case of the energy domain, SimModel is a
simulation data model which is tailored towards en-
ergy-related data required by most popular energy
simulation tools. In other domains, such as the eco-
nomic one, it is still not clear that there is such rep-
resentative model.
One of the conclusions that can be highlighted in
the research outlined in this paper is that even with
more flexibility to generate the simulation models
from the data models to perform simulations for
each tools, ad-hoc adapters need to be developed to
provide this interoperability.
The next steps in this research are to further de-
velop the ontology-based solution adopted in the
DDM to carry out fully-automated energy simula-
tions with EnergyPlus and CitySIM.
6 ACKNOWLEDGEMENTS
Part of the work presented in this paper is based on
research conducted within the project “Optimised
Energy Efficient Design Platform for Refurbishment
at District Level”, which has received funding from
the European Union Horizon 2020 Framework Pro-
gramme (H2020/2014-2020) under grant agreement
n° 680676.
REFERENCES
Amirebrahimi, S., Rajabifard, A., Mendis, P., & Ngo, T. 2015.
A Data Model for Integrating GIS and BIM for Assessment
and 3D Visualisation of Flood Damage to Building. Locate,
15, 10-12.
Beetz, J., Van Leeuwen, J., & De Vries, B. 2009. IfcOWL: A
case of transforming EXPRESS schemas into ontologies.
Artificial Intelligence for Engineering Design, Analysis and
Manufacturing, 23(01), 89-101.
Bontas, E. P., Mochol, M., & Tolksdorf, R. 2005. Case studies
on ontology reuse. In Proceedings of the IKNOW05 Inter-
national Conference on Knowledge Management (Vol. 74).
BuildingSMART. (2015). IFC4 Release Summary.
http://www.buildingsmart-tech.org/specifications/ifc-
releases/ifc4-release
Crapo, Andrew, Ray Piasecki, and Xiaofeng Wang. 2011. The
smart grid as a semantically enabled internet of things. In
Implementing Interoperability Advancing Smart Grid
Standards, Architecture and Community: Information In-
teroperability Track, Grid-Interop Forum (Vol. 2011).
Döllner, J., & Hagedorn, B. 2007. Integrating urban GIS, CAD,
and BIM data by service based virtual 3D city models. R. et
al.(Ed.), Urban and Regional Data Management-Annual,
157-160.
Gröger, G., & Plümer, L. 2012. CityGML–Interoperable se-
mantic 3D city models. ISPRS Journal of Photogrammetry
and Remote Sensing, 71, 12-33.
Gröger G, Kolbe T H, Nagel C, Häfele K H. 2012. OGC City
Geography Markup Language (CityGML) Encoding Stand-
ard Version 2.0. http://www.opengis.net/spec/citygml/2.0
ISO 16739. 2013 Industry Foundation Classes (IFC) for data
sharing in the construction and facility management indus-
tries.
Kolbe, T. H. 2009. Representing and exchanging 3D city mod-
els with CityGML. In 3D geo-information sciences (pp. 15-
31). Springer Berlin Heidelberg.
Laat, R., and Berlo, L. 2011. Integration of BIM and GIS: The
Development of the CityGML GeoBIM Extension. In Ad-
vances in 3D Geo-Information Sciences, Berlin, Germany,
211-225.
Lilis, G. N., Giannakis, G., & Rovas, D. 2015. Detection and
Semi-automatic correction of geometric inaccuracies in IFC
files. IBSPA Conference, Hyderabad, India.
Lilis, G.N., Giannakis, G.I., Katsigarakis, K., Rovas, D.V.,
Costa, G., Sicilia, A., & Garcia, M.A. 2016. Simulation
model generation combining IFC and CityGML data. In
Proceedings of the 11th European Conference on Product
& Process Modelling. Limassol, Cyprus, pp. XXX-XXX.
Mao, B., & Ban, Y. 2011. Online Visualization of 3D City
Model Using CityGML and X3DOM. Cartographica: The
International Journal for Geographic Information and Ge-
ovisualization, 46(2), 109-114.
Métral, C., Billen, R., Cutting-Decelle, A. F., & Van
Ruymbeke, M. 2010. Ontology-based approaches for im-
proving the interoperability between 3D urban models.
Journal of Information Technology in Construction, 15,
169-184.
O'Donnell, J., See, R., Rose, C., Maile, T., Bazjanac, V., &
Haves, P. 2011. SimModel: A domain data model for whole
building energy simulation. Proceedings of Building Simu-
lation 2011: 12th Conference of International Building
Performance Simulation Association, Sydney.
Paslaru, E. 2005. Using context information to improve ontolo-
gy reuse. In Doctoral Workshop at the 17th Conference on
Advanced Information Systems Engineering CAiSE (Vol.
5).
Pauwels, P., Corry, E., & O’Donnell, J. 2014. Representing
SimModel in the web ontology language. In American Soci-
ety of Civil Engineers (pp. 2271-2278).
Pauwels, P., & Terkaj, W. (2016). EXPRESS to OWL for con-
struction industry: towards a recommendable and usable
ifcOWL ontology. Automation in Construction, 63, 100-
133.
Romero, A.; Egusquiza, A.; Izkara, J.L. 2014. Integrated deci-
sion support tool in energy retrofitting projects for sustaina-
ble urban districts. In proceedings of the SB14 World Sus-
tainable Buildings. Barcelona. October 2014.
Schevers, H., & Drogemuller, R. 2005. Converting the industry
foundation classes to the web ontology language. In Seman-
tics, Knowledge and Grid, 2005. SKG'05. First Interna-
tional Conference on (pp. 73-73). IEEE
Sicilia, A., Madrazo, L., & Pleguezuelos J. 2014. Integrating
multiple data sources, domains and tools in urban energy
models using semantic technologies. In Proceedings of the
10th European Conference on Product & Process Model-
ling. Vienna, Austria, pp. 837-844.
Simperl, E. 2009. Reusing ontologies on the Semantic Web: A
feasibility study. Data & Knowledge Engineering, 68(10),
905-925.
Succar, B. 2009. Building information modelling framework: A
research and delivery foundation for industry stakeholders.
Automation in Construction, 18(3), 357-375.
Vimmr, T., Loga, T., Diefenbach, N., Brita, S., & Lucie, B.
2013. Tabula - Residential Building Typologies. In 12th
European Countries - Good practice example from the
Czech republic. CESB13 Prague, Czech Republic.